针对传统粒子群算法(PSO)中存在的易陷入局部最优解和后期收敛速度慢的问题,首次提出一种新混合粒子群算法(NHPSO),采用杂交粒子群算法和固定惯性权重策略,并把简化的二次插值法融入杂交粒子群算法中。实验证明新算法大大提高了收敛速度,改善了解的质量。对阵列天线特殊主瓣形式的波束赋形和旁瓣电平优化结果取得了非常好的效果,计算机仿真证实该新算法应用于此类问题非常有效。
A hybrid Particle Swarm Optimization(PSO) algorithm is proposed with fixed inertia weight in the hybrid particle swarm optimization algorithm,and a simplified quadratic interpolation method is integrated into this algorithm,aiming at overcoming easily trapping in the local extreme points and slow evolving speed of convergence. The experiment shows that this new algorithm improved the global search ability and the quality of optima. The results of both mainlobe shaping and sidelobe levels are very effective. The simulation results prove that the proposed hybrid new algorithm is efficient in this kind of problems.